/
event.go
371 lines (313 loc) · 8.82 KB
/
event.go
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
// Copyright The OpenTelemetry Authors
// SPDX-License-Identifier: Apache-2.0
package groupbytraceprocessor // import "github.com/open-telemetry/opentelemetry-collector-contrib/processor/groupbytraceprocessor"
import (
"context"
"errors"
"fmt"
"hash/maphash"
"sync"
"time"
"go.opencensus.io/stats"
"go.opencensus.io/tag"
"go.opentelemetry.io/collector/pdata/pcommon"
"go.opentelemetry.io/collector/pdata/ptrace"
"go.uber.org/zap"
)
const (
// traces received from the previous processors
traceReceived eventType = iota
// traceID to be released
traceExpired
// released traces
traceReleased
// traceID to be removed
traceRemoved
)
var (
errNoTraceID = errors.New("trace doesn't have traceID")
seed = maphash.MakeSeed()
hashPool = sync.Pool{
New: func() any {
var hash maphash.Hash
hash.SetSeed(seed)
return &hash
},
}
eventTagKey = tag.MustNewKey("event")
)
type eventType int
type event struct {
typ eventType
payload any
}
type tracesWithID struct {
id pcommon.TraceID
td ptrace.Traces
}
// eventMachine is a machine that accepts events in a typically non-blocking manner,
// processing the events serially per worker scope, to ensure that data at the consumer is consistent.
// Just like the machine itself is non-blocking, consumers are expected to also not block
// on the callbacks, otherwise, events might pile up. When enough events are piled up, firing an
// event will block until enough capacity is available to accept the events.
type eventMachine struct {
workers []*eventMachineWorker
close chan struct{}
metricsCollectionInterval time.Duration
shutdownTimeout time.Duration
logger *zap.Logger
onTraceReceived func(td tracesWithID, worker *eventMachineWorker) error
onTraceExpired func(traceID pcommon.TraceID, worker *eventMachineWorker) error
onTraceReleased func(rss []ptrace.ResourceSpans) error
onTraceRemoved func(traceID pcommon.TraceID) error
onError func(event)
// shutdown sync
shutdownLock *sync.RWMutex
closed bool
}
func newEventMachine(logger *zap.Logger, bufferSize int, numWorkers int, numTraces int) *eventMachine {
em := &eventMachine{
logger: logger,
workers: make([]*eventMachineWorker, numWorkers),
close: make(chan struct{}),
shutdownLock: &sync.RWMutex{},
metricsCollectionInterval: time.Second,
shutdownTimeout: 10 * time.Second,
}
for i := range em.workers {
em.workers[i] = &eventMachineWorker{
machine: em,
buffer: newRingBuffer(numTraces / numWorkers),
events: make(chan event, bufferSize/numWorkers),
}
}
return em
}
func (em *eventMachine) startInBackground() {
em.startWorkers()
go em.periodicMetrics()
}
func (em *eventMachine) numEvents() int {
var result int
for _, worker := range em.workers {
result += len(worker.events)
}
return result
}
func (em *eventMachine) periodicMetrics() {
numEvents := em.numEvents()
em.logger.Debug("recording current state of the queue", zap.Int("num-events", numEvents))
stats.Record(context.Background(), mNumEventsInQueue.M(int64(numEvents)))
em.shutdownLock.RLock()
closed := em.closed
em.shutdownLock.RUnlock()
if closed {
return
}
time.AfterFunc(em.metricsCollectionInterval, func() {
em.periodicMetrics()
})
}
func (em *eventMachine) startWorkers() {
for _, worker := range em.workers {
go worker.start()
}
}
func (em *eventMachine) handleEvent(e event, w *eventMachineWorker) {
switch e.typ {
case traceReceived:
if em.onTraceReceived == nil {
em.logger.Debug("onTraceReceived not set, skipping event")
em.callOnError(e)
return
}
payload, ok := e.payload.(tracesWithID)
if !ok {
// the payload had an unexpected type!
em.callOnError(e)
return
}
em.handleEventWithObservability("onTraceReceived", func() error {
return em.onTraceReceived(payload, w)
})
case traceExpired:
if em.onTraceExpired == nil {
em.logger.Debug("onTraceExpired not set, skipping event")
em.callOnError(e)
return
}
payload, ok := e.payload.(pcommon.TraceID)
if !ok {
// the payload had an unexpected type!
em.callOnError(e)
return
}
em.handleEventWithObservability("onTraceExpired", func() error {
return em.onTraceExpired(payload, w)
})
case traceReleased:
if em.onTraceReleased == nil {
em.logger.Debug("onTraceReleased not set, skipping event")
em.callOnError(e)
return
}
payload, ok := e.payload.([]ptrace.ResourceSpans)
if !ok {
// the payload had an unexpected type!
em.callOnError(e)
return
}
em.handleEventWithObservability("onTraceReleased", func() error {
return em.onTraceReleased(payload)
})
case traceRemoved:
if em.onTraceRemoved == nil {
em.logger.Debug("onTraceRemoved not set, skipping event")
em.callOnError(e)
return
}
payload, ok := e.payload.(pcommon.TraceID)
if !ok {
// the payload had an unexpected type!
em.callOnError(e)
return
}
em.handleEventWithObservability("onTraceRemoved", func() error {
return em.onTraceRemoved(payload)
})
default:
em.logger.Info("unknown event type", zap.Any("event", e.typ))
em.callOnError(e)
return
}
}
// consume takes a single trace and routes it to one of the workers.
func (em *eventMachine) consume(td ptrace.Traces) error {
traceID, err := getTraceID(td)
if err != nil {
return fmt.Errorf("eventmachine consume failed: %w", err)
}
var bucket uint64
if len(em.workers) != 1 {
bucket = workerIndexForTraceID(traceID, len(em.workers))
}
em.logger.Debug("scheduled trace to worker", zap.Uint64("id", bucket))
em.workers[bucket].fire(event{
typ: traceReceived,
payload: tracesWithID{id: traceID, td: td},
})
return nil
}
func workerIndexForTraceID(traceID pcommon.TraceID, numWorkers int) uint64 {
hash := hashPool.Get().(*maphash.Hash)
defer func() {
hash.Reset()
hashPool.Put(hash)
}()
_, _ = hash.Write(traceID[:])
return hash.Sum64() % uint64(numWorkers)
}
func (em *eventMachine) shutdown() {
em.logger.Info("shutting down the event manager", zap.Int("pending-events", em.numEvents()))
em.shutdownLock.Lock()
em.closed = true
em.shutdownLock.Unlock()
done := make(chan struct{})
// we never return an error here
ok, _ := doWithTimeout(em.shutdownTimeout, func() error {
for {
if em.numEvents() == 0 {
return nil
}
time.Sleep(100 * time.Millisecond)
// Do not leak goroutine
select {
case <-done:
return nil
default:
}
}
})
close(done)
if !ok {
em.logger.Info("forcing the shutdown of the event manager", zap.Int("pending-events", em.numEvents()))
}
close(em.close)
}
func (em *eventMachine) callOnError(e event) {
if em.onError != nil {
em.onError(e)
}
}
// handleEventWithObservability uses the given function to process and event,
// recording the event's latency and timing out if it doesn't finish within a reasonable duration
func (em *eventMachine) handleEventWithObservability(event string, do func() error) {
start := time.Now()
succeeded, err := doWithTimeout(time.Second, do)
duration := time.Since(start)
_ = stats.RecordWithTags(context.Background(), []tag.Mutator{tag.Upsert(eventTagKey, event)}, mEventLatency.M(duration.Milliseconds()))
if err != nil {
em.logger.Error("failed to process event", zap.Error(err), zap.String("event", event))
}
if succeeded {
em.logger.Debug("event finished", zap.String("event", event))
} else {
em.logger.Debug("event aborted", zap.String("event", event))
}
}
type eventMachineWorker struct {
machine *eventMachine
// the ring buffer holds the IDs for all the in-flight traces
buffer *ringBuffer
events chan event
}
func (w *eventMachineWorker) start() {
for {
select {
case e := <-w.events:
w.machine.handleEvent(e, w)
case <-w.machine.close:
return
}
}
}
func (w *eventMachineWorker) fire(events ...event) {
w.machine.shutdownLock.RLock()
defer w.machine.shutdownLock.RUnlock()
// we are not accepting new events
if w.machine.closed {
return
}
for _, e := range events {
w.events <- e
}
}
// doWithTimeout wraps a function in a timeout, returning whether it succeeded before timing out.
// If the function returns an error within the timeout, it's considered as succeeded and the error will be returned back to the caller.
func doWithTimeout(timeout time.Duration, do func() error) (bool, error) {
done := make(chan error, 1)
go func() {
done <- do()
}()
select {
case <-time.After(timeout):
return false, nil
case err := <-done:
return true, err
}
}
func getTraceID(td ptrace.Traces) (pcommon.TraceID, error) {
rss := td.ResourceSpans()
if rss.Len() == 0 {
return pcommon.NewTraceIDEmpty(), errNoTraceID
}
ilss := rss.At(0).ScopeSpans()
if ilss.Len() == 0 {
return pcommon.NewTraceIDEmpty(), errNoTraceID
}
spans := ilss.At(0).Spans()
if spans.Len() == 0 {
return pcommon.NewTraceIDEmpty(), errNoTraceID
}
return spans.At(0).TraceID(), nil
}